Robust audio watermarking based on transform domain and SVD with compressive sampling framework

Abstract

The growth of the internet and digital data has resulted forgery, modification and sharing of digital data without property rights. Audio watermarking is one of a solution to protect the copyright of an audio from copyright infringement. This paper proposes an audio watermarking method which is robust against attacks and high capacity. First, a synchronization bit is added to the audio host. After the audio host is decomposed by Lifting Wavelet Transform (LWT), then choose a subband from the output of LWT to be transformed by discrete cosine transform (DCT). Next, the matrix of the signal from DCT is selected for the singular value decomposition (SVD) process, so that is obtained U, S and V matrix. S matrix is embedded with the watermark. Before the embedding process, the watermark image is compressed by Compressive Sampling. The results show that the proposed watermarking system is highly robust against a kind attack of LPF, resampling, and linear speed change which is proven by its BER is zero.